Effects of transport stress on fecal microbiota in healthy donkeys

Background: With the development of large-scale donkey farming in China, long-distance transportation has become a common practice, and the incidence of intestinal diseases after transportation has increased. Intestinal microbiota is important for health and disease, and whether transportation disturbs donkey intestinal microbiota has not been investigated. This study aims to determine the effects of transportation on the fecal microbiota of healthy donkeys using 16S rDNA sequencing. Results: Fecal samples were collected from the rectum of 12 Dezhou donkeys before and after transportation. Results show that long-distance transportation can induce severe stress in donkeys and result in significantly lower level of bacterial richness index compared with that before transport (p=0.042) without distinct changes in diversity. This marked decrease in specific bacterial richness, such as for Eubacterium, Streptococcus, and Coriobacteriaceae, might be associated with the restricted synthesis of anti-inflammatory cytokines and metabolites, such as short chain fatty acids (SCFAs) that potentially contribute to disease development after the transport. Conclusions: Further studies are required to understand the potential effect of these microbiota changes on the development of donkey intestinal diseases. Preventative and therapeutic measures for donkeys before and after transportation should focus on providing diverse and rich bacterial microbiota and probiotic flora.

microbial communities can be associated with a wide range of diseases in equine gastrointestinal tract, including colitis [10], laminitis [11], equine grass sickness [12], and transient diarrhea in foals [13]. Administration of medications, transportation, fasting, and abrupt changes in diet could induce changes on horse intestinal microbiota [14][15][16]. However, the influence of these factors on donkeys has been poorly studied. Culture-and isolate-based conventional methods used to study microbiota have limited sensitivity for the assessment of complex microbial populations; for example, only less than 1% of microbes in any marine habitats can be cultured under standard laboratory conditions [17]. A recent next-generation sequencing and bioinformatics based technology allows for a detailed quantitative analysis of the microbiota and could be used to understand the influence of specific factors, such as antimicrobial treatment, diarrhea, and mild asthma, on horse intestinal or respiratory tract microbiota [18,19].
Transporting donkeys from traditional donkey-concentrated areas for fattening and breeding has become a major breeding model in China and has been accompanied by the increase in long-duration transportation. Our previous statistics show that long distance transportation led to high morbidity and mortality among donkey during recovery period, and intestinal diseases are one of the main diseases that gradually became one of the key factors restricting the development of donkey breeding industry in China. Despite the high damage of donkey transportation, the effect of transportation on donkey gut microbiota is poorly understood. For the first time, we used highthroughput pyrosequencing to evaluate the effects of transport on donkey fecal microbiota, understand the pathophysiology of diseases related to gut microbiota, and develop effective preventive or treatment options.

Transportation of donkey alters its hormonal levels
The concentration of serum Cortisol hormone (Cor), heat-shock protein 90 (HSP90), and adrenocorticotrophic hormone (ACTH) significantly differed between before and immediately after transportation. Serum ACTH, Cor, and HSP90 of the donkeys were significantly increased (p<0. 05) on the day of arrival compared with those on the day before transportation (Fig. 1).
Sequencing quality data and alpha diversity analysis 5 The microbiota composition of the fecal samples was assessed by sequencing the bacterial 16S rRNA V3 + V4 region. A total of 1,233,776 pairs of reads were obtained from the 24 samples that were sequenced. Double-end read splicing and filtering resulted in 1,075,841 clean tags, and each sample produced 44,827 clean tags on average. With the use of QIIME (version 1.8.0) UCLUST software based on 97% sequence similarity, the tags were clustered into OTUs. The number of OTUs in fecal samples after transportation (AF1-12) decreased markedly relative to that before transportation (BF1-12) ( Fig.   2A). The Venn diagram of OTUs was illustrated (Fig. 2B) and the Chao, Simpson, and Shannon indexes were calculated (Fig. 2C). A significant decrease in Chao index was found in the fecal samples after transportation (AF group) compared with that before transportation (BF group). No other differences in alpha diversity were identified.
Beta diversity analysis Differential analysis, including PCoA (Fig. 3A) and UPGMA, between groups based on the weighted UniFrac algorithm were performed to further explore the relationship between different bacterial communities before and after transport (Fig. 3B). Fig. 3A shows that the percentage of variation explained by PC1 and PC2 were 43.01% and 17.01%, respectively. However, distinct clusters were not visually evident with PCoA, indicating no significant difference in the bacterial composition before and after transport. The phylogenetic tree based on the weighted UniFrac also revealed that the differences in bacterial community were difficult to discern visually, and this finding was in accordance with the PCoA result.

Phylogenetic analysis
The relative bacteria abundance of phylum in each sample (Fig. 4A) and group (Fig. 4B) were drawn using R language tools. As shown in Fig. 4B, 12 phyla had a mean relative abundance of more than 1%. The predominant phyla in each group were Bacteriodetes, Firmicutes, Proteobacteria, Verrucomicrobia, Fusobacteria, Fibrobacteres, and Spirochaetes. The phylum-level analysis also revealed that the most prevalent phyla of the microbial community after transport were similar to those before transport. However, transport affected the relative abundance of phyla; for example, an increased relative abundance of Bacteroidetes (median: BF 32.6% to AF 40.9%) and Firmicutes (median: BF 23.3% to AF 30.5%) and a decreased relative abundance of Proteobacteria (median: BF 18.1% to AF 10.3%) and Verrucomicrobia (median: BF 9.2% to AF 5.8%) were observed after transport.
The line discriminant analysis (LDA)-effect size (LEfSe) method was used for the quantitative analysis of biomarkers in the microbiota among each group. The LDA score was set at 2.0, and different genera with LDA threshold >2 were considered significant biomarkers. The cladogram is shown in Fig.   5A, and the LDA score distribution map is shown in Fig. 5B. Several genera were more abundant in BF samples than in AF samples according to LEfSe analysis. Eubacterium genus, Coriobacteriaceae family, Streptococcus species, Atopostipes,and Pseudomonas genera were enriched in BF samples compared with those in AF samples.

Discussion
The bacterial microbiota of foals plays an important role in various digestive tract diseases, and early colonization and development is a dynamic process. However, donkey bacterial microbiota is poorly studied. Initial culture-based studies only focus on a narrow spectrum of the fecal bacterial microbiota, and subsequent culture-independent studies based on 16S ribosomal RNA (rRNA) gene sequencing studies have provided insights into the rapid development of bacterial microbiota. In this study, the effect of transport on the phylogenetic composition of donkey fecal microbiota was analyzed. Differences in the relative abundances of phyla, classes, and orders and the loss of bacterial diversity and richness were observed.
ACTH and Cor levels increase under stress to deal with changes in the external environment. These hormones are important indexes in the stress reaction of animals, including beef cattle, piglets, chicken, and horses. During stressful situations, such as transportation, the ACTH and cortisol content in plasma increases variably [20][21][22]. HSP90 is an important stress protein in organisms because it is rapidly activated and synthesized during stress reaction [23]. In this study, the ACTH, Cor, and HSP90 levels significantly increased (p 0.05) after transport, and this finding is in agreement with other studies worldwide. Therefore, cold weather, crowded environment, bumpy transportation, and changes in environment and feeding patterns after arrival cause severe stress to the donkeys. Transport stress rapidly affects the composition of gut microbiota and host physiology through the generation of bioactive metabolites [14,24]. Our study revealed that donkey fecal microbiota after transport had significant decreases in specific bacterial richness compared with the controls, and this finding may be related to gastrointestinal diseases. LEfSe analysis revealed that after transport, significant decreases were found for the short chain fatty acids (SCFAs)-producing bacteria, including Eubacterium genus and Coriobacteriaceae family [25]. SCFAs are not included the diet but synthesized by colonic commensal bacteria from dietary carbohydrate; these substances plays a key role in the energy metabolism of the colonic epithelial cells and is important in the maintenance of colon health in humans [26]. SCFAs also act on immune cells, such as mononuclear phagocytes and lymphocytes, and participate in intestinal immune regulation by influencing the release of inflammatory factors and chemotaxis, thereby playing an important role in intestinal defense against pathogenic bacteria [27].
Interfering with the SCFAs synthesis in the colon may result in diarrhea; an increased production of SCFAs enhances the colonic fluid production and corrects the dehydration associated with acute diarrhea [28]. Moreover, the levels of butyrate produced by Eubacterium were significantly lower during and immediately after diarrhea than during a diarrhea-free period of normal health [29]. The number of Coriobacteriaceae family, which is found abundant in healthy gut, was significantly decreased in microscopic colitis cases [30]. The marked decrease in the relative abundance of these SCFAs-producing bacteria might therefore reflect that transport stress can interfere with SCFAs synthesis and may be important in the pathophysiology. Hence, donkeys are at a high risk of diarrhea after transport.
In addition to the marked decrease in SCFAs-producing bacteria, significantly low rate of Streptococcus, Atopostipes,and Pseudomonaswas observed after transport. Atopostipes and Pseudomonas are bacteria producing branched chain fatty acids (BCFAs). Atopostipes is a bacterial genus of Lactobacilliales order under the Bacilli class, a gram-positive bacterium. As a lactic acid bacteria, Atopostipes has a strong positive correlation with BCFAs by metabolizing valine and tryptophan to BCFAs [31]. Pseudomonas produce BCFAs through the deamination of branched chain amino acids such as leucine, isoleucine, and valine [31]. BCFAs, which are primarily saturated fatty acids (FA), are normal constituents in the gut throughout the human life cycle. Similar to SCFAs, BCFAs also have an important influence on intestinal health and are related to various health conditions. These compounds are metabolized by enterocytes and have a beneficial role against inflammation in the premature intestine, alter the microbiota, and increase the expression of antiinflammatory cytokines [32]. Streptococcus is an enriched taxon according to LEfSe in the healthy horses [33]. Streptococcus and Lactococcus are negatively correlated with inflammatory parameters (TNF-α, LPS, and H 2 O 2 yield) [34]. A probiotic combination containing Streptococcus thermophilus protected the bowel and improved colon inflammation in experimentally induced inflammatory bowel disease in rats [35].These results indicate that transport stress significantly reduces the number of probiotics such as lactic acid bacteria and might disturb the synthesis of BCFAs, thereby increasing the rate of inflammatory bowel disease.
Previous studies have found that potential confounding factors, such as age [33], gender [36], diets [37] and environmental change [38] could impact the bacterial microbiota of differnet body sites.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests. The experiment was conducted on 12 male Dezhou donkeys aged 10-12 months weighing (140.8±5.2 kg, mean±SD). The donkeys were clinically healthy and provided free access to water and feed composed of straw and commercial concentrates daily. All donkeys had no previous experience of road transport. During transportation, the average environmental temperature and humidity were −10 °C and 28%, respectively. The surrounding walls of the truck (13.4 m long and 5.6 m wide) were equipped with iron guardrails, and the floor was iron with extremely thin bedding materials. The truck did not have roof coverings, and the donkeys were therefore exposed to different weather conditions.

Bioinformatics and data analysis
The raw paired-end reads from the original DNA fragments were merged using FLASH32 and assigned to each sample according to the unique barcodes. QIIME (version 1.8.0) UCLUST software was used based on 97% sequence similarity, and the tags were clustered into operational taxonomic units    Composition and relative abundance of dominant phyla in the bacterial communities of different samples (a) and groups (b) before and after transport. Other: Bacterial taxa with ≤1% abundance, Unknow: Sequences which could not be calssified.

Figure 5
Linear discriminant analysis identifying differences between BF and AF group. a Phylogenetic profile of specific bacterial taxa and predominant bacteria among the two different groups. b LDA score of 2.0 were used as thresholds for significance in LEfSe analyses.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.
ARRIVE checklist.pdf